OK, enough conversation derailing, please. These have been the rules pretty much the whole time, see FAQ - discuss.pixls.us
3 Likes
I think it comes pretty close to the mood that the linked preset tried to offer, most other edits fall in the warm tones. Those results are pleasant, but they don’t really look like the end goal. I really like your take because it’s the most faithful to the example provided. In the end, it all comes down to personal preference and how much one wants to stick with the objective.
3 Likes
Yes of course. You’re right. It’s a matter of taste. And although I admit that I like the style, the goal here ultimately was to learn.
Tinkering with DT to get to that result has taught me so much more than just trying to understand it module by module.
I think now I have a “solution” that I can work with and base other edits of of it. To me the most important part was to learn which modules do the stuff that I did with a simpler slider when using LR. I’m still nowhere near fully understanding DT but it’s gotten a lot easier. I think I will post my solution next to how the preset does it so others can maybe also relate to it more.
And now that I got a little more experience with DT, I have a general opinion - and I hope you guys won’t throw stones at me for saying it 
DT is an amazing piece of software which gives you a bunch of control. But I feel in terms of user friendliness and overall understandability it lacks a lot. It feels like it’s a program that has been written and designed by coders, mathematicians, maybe even one or the other physicist that do know how photo editing works on a technical level, but didn’t manage to make the tool user friendly.
It gives you so many options and sliders and checkboxes and modules that one has to feel overwhelmed by them - at least I did. I believe that DT could be simpler. More intuitiv. Tidier. I’m no programmer whatsoever so I can’t speak to how well DT is coded. I’m just a photographer and designer. A visual person. As such, I rely on a clean interface that lets me figure out quickly, how to do stuff. I don’t want to have to read through pages of manuals just to get a thing done that should take a minute or less.
I understand that it is an open source project. I also understand that probably noone gets paid to contribute to the program.
I just wish that some day in the future someone with the ability could make it a more user friendly tool and in terms of the UI y do have a ton of ideas that would make using DT easier or more understandable. I also have ideas about how to maybe make it seem less confusing. If there is already any topic about that, I’m happy to add my 5 cents to it. 
But enough complaining. Thanks everyone for this fun edit. Later today imma post the final comparison with the module overview and that should conclude this post. But I already have a new style in mind that I want to try. 
4 Likes
I totally get where you’re coming from, I’ve always been a lightroom aficionado and just recently switched to Affinity Suite and found interest in darktable.
I’ve seen the improvements you made post after post, and it’s quite encouraging.
I was in the same situation some months ago, I always heard of darktable but was intimidated because it gives you many tools but of course there’s always the risk of derailing because the modules can “backfire” if you don’t properly instruct yourself. After a while it clicks, tho.
I also really enjoy the fact that you developed and documented your journey during this PlayRaw. It’s truly inspiring to see fresh and different approaches so anyone can learn.
1 Like
It would be interesting for example to analyse the database of a proficient user or users of DT say maybe from the last 2 or three years as it has matured. I suspect if you did a module count on the type and number of modules used you would reveal to new users that DT edit usual require a handful of modules often used more than one time per edit and with masking to perform the majority of edits… The rest of the modules are there for special use cases. The devs have tried to provide a streamlined selection with a module group preset list but I suspect many new users have no idea about this. It took some time for me to become infinitely more streamlined in module selection and use but it came. I now use a relatively small group of modules with some go to presets that can be tweaked. When I first started I would have these long history stacks throwing modules at the image to try and get a good look… When I come across one of those old edits… I laugh at myself…
2 Likes
I’m at the point I’m already embarrassed to go back to files after a week of developing them, just because I learn so much from all the precious input of the community around this software 
By the way, I wanted to experiment a little and followed the video from the website which was linked to edit the picture like it’s done there and this is what I got:
- Downloaded a Hald8 .png file (maybe even a Hald5 would have done the job, but I’m used to the 64х64х64)
- Applied colour and light edits to the hald (no clarity, nor details or noise reduction: we can’t mess with the hald, they just corrupt the properties of the LUT)
- Exported as a 16bit .tif sRGB file from Lightroom Mobile
- Converted the .tif to .png since it can’t be done from LR app anymore
- Then I could have converted the .png to .cube but since darktable can read the png I didn’t bother doing the conversion so
- Directly loaded the .png with the LUT 3D module in dt and here’s the result of that preset applied to the provided image.
It doesn’t translate really well, but it’s a starting point, I guess?
In the end we’re back to what @paperdigits initially suggested: desaturate most colors except the one you want to bring out, in fact the Hald8 shows us this:
Honestly - I did not understand the process you just explained to get the LUT hahaha but I think that would be too easy. It’s not just “desaturate everything except oranges” and you’re done. It’s the intricate adjustment of a couple of colors to “lean” towards the main color (in this example orange), adjust their luminosity, desaturate colors that are unwanted (blue to magenta) and then adjust shadows, highlights, curve, black and white points etc. I’m not even sure a LUT could do that but again…I don’t know much about how LUTS work and operate. I just know they are applied and give a certain look. I know - I’m ignorant in that sense.
Well but now for the final reveal. This is my before and after:
And this is how I got there (left col LR steps, right col DT modules):
1. Exposure -0.3 Exposure Module
2. Contrast 70 Color Balance RGB Module
3. Highlights -40 Shadows Highlights Module
4. Shadows +25 Shadows Highlights Module
5. Whites -70 Shadows Highlights Module
6. Blacks +20 Exposure Module
7. Curve lift blacks J Tone Curve Module
8. Vibrance +30 Color Balance RGB Module
9. Saturation +16 Color Balance RGB Module
10. HSL Color Equalizer Module
RED H 0 S -20 L -10
ORANGE H -10 S -30 L +10
YELLOW H -100 S +30 L +30
GREEN H 0 S -100 L 0
AQUA H 0 S -100 L 0
BLUE H 0 S -100 L 0
PURPLE H 0 S -100 L 0
MAGENTA H 0 S -100 L 0
11. Texture +15 Local Contrast Module
12. Clarity +50 Contrast Equalizer Module
13. Dehaze +40 Haze Removal Module
14. Vignette -40 Feather 80 Vignetting Module
15. Splittone Split Toning Module
Highlights H 220 Saturation 10
16. Noise Redutcion +15 Denoise (profiled) Module
17. Color Noise +35 Denoise (profiled) Module
And this is the Style exported:
UFP _ BLACK ORANGE.dtstyle (7.6 KB)
1 Like
You can also use channel mixer to get the base for the orange. I made an episode about it.
This is what the photo looks like after all the corrections and white balancing:
With channel mixer in a new instance of color calibration we reduce the amount of green to make yellow more orange and increase the green in the blue channel to protect the white point. We call the instance “yellow to orange”:
In the same instance, we also reduce the brightness of the green channel and increase it for the red channel so that the image as a whole becomes darker and the orange tones brighter:
Now we use color equalizer to desaturate everything except warm tones:
In the same instance, we can make yellowish orange tones even more orange:
With color balance rgb we can increase the contrasts:
And in the sigmoid module increase the tarrget black to increase the black point:
At the end, you can adjust what should be orange by playing with all three sliders in the “yellow to orange” instance in the red channel. In this case, I reduce the orange in the buildings a little:
_UFP0298.NEF.xmp (14,2 KB)
5 Likes
You load the Hald8 .png in Lightroom, copy the edits from “light” and “color” tabs in the other image, paste them on the Hald8 used for the LUT, export to TIF/16bit (you need to convert it to PNG/16bit on the computer), and you get a LUT to feed to the program you choose to replicate edits in.
That’s why I wanted to show that the “preset” is not a “one size fits all” type of thing, as you figured out by developing your own workflow in darktable. You want the look of it, I know but in this case it really is
Thanks to the LUT, I’ve been able to show what this image paired with that preset translates to in other programs, it’s the base from where you start making edits. I’ll try to explain: it replicates color and light changes made from a program to another one, it really just comes down to “desaturate this, saturate that, lighten this, darken that”, then paste those to the hald .png and voilà. Any software you will input that hald into will use it as a “Look Up Table” and convert your image based on what you did in the other one, be it PS, be it AfPhoto or anyone that accepts LUT files.
The result I attached objectively is what has been altered in the video to make that specific image look like that. The YT channel involved specifically hides passwords for the presets just to make people’s life harder in order to farm engagement and views on YouTube in hope that people will watch it enough time to turn a profit.
Let me be clear: the latest image is not how I would edit. I already made my PlayRaw.
That’s what the edits in LR would give and what works with the “demo image”. As I said, it’s the starting point, which is specifically why I’m not a huge fan of presets salesmen (except for trusted color industry companies and movie professionals) because they sell you any preset after showing it on cherry-picked pictures.
The analogy is this: It’s like I was trying to convince you to use my sidecar .xmp, asked for a tip, but then you didn’t like it, and you wasted money and time, you bought air.
The things that can’t transfer with a LUT are “clarity, dehaze, noise reduction, color noise reduction, sharpening” and of course no lens correction nor chromatic aberration fixes because they are properties that just mess up the hald and don’t get translated with lookup tables. That’s up to user to replicate in other software because they are not the same. Ideally, one wouldn’t even change the light if you just need to translate colors and light will be handled later.
In the end if you try to compare programs too much you’ll never be happy with the result because tools might have the same name but behave differently, the math differs and as you can see I went through a lot of work just to make a picture with that exact preset, and it wasn’t liked (which I already expected 
) because it comes down to what you feel like doing on your image. You know best what you want to do with that picture, and the preset is based on edits made in another software which handles images in its own way by another person. It works on the image that it was developed on, and most certainly is just a starting point for any other picture. That’s why learning through these .XMPs is really valuable.
I hope it’s clearer, cheers and have fun with dt! Looking forward to another PlayRaw.
1 Like
For step 2 the grey fulcrum slider in the mask tab of rgbCB could be useful to tweak your look… it is the set point (pivot) for what gets made darker and lighter when you use contrast…it might be a nice way to fine tune that aspect of your edit…
1 Like
Oh wow. Thanks for the detailed explanation. Channel mixer and sigmoid are still two modules I haven’t done much with so I’ll check it out!
@DaDaLand I appreciate your explanation. It made it a bit clearer but not much. The thing I’m getting at is: no matter how you achieve a look that you’re after, there’s lessons to be learned. And from not knowing even where to start in DT to now being able to pick and choose between various modules to achieve something, to me is a great step up. Let’s see what I learn in the next one. I already posted it. It’s called moody greens of you wanna check it out
Yeah, this software really grows on you and each time experimenting is key, I always try to change the modules I use.
In my very first reply, I just tried to minimize the modules involved and used the recently learnt “colour equalizer” as seen in @s7habo’s most recent video (or at least that’s what I tried to go for after acquiring some of his knowledge 
). Keep an eye on those videos, they are very inspiring and resourceful.
1 Like
Totally. That’s how I learned most of what the modules do for this edit. There’s only a few users out there who make DT tutorials and some resonate to me more than others but there’s always something useful.
What @DaDaLand is referring to is explained here for Rawtherapee… but basically you can load an identity png in almost any program and copy your style, edits, or if in LR then a preset and then save that as a png to be uses as a lut in other programs…it can be a convenient way or at least one strategy to get to a starting point for a creative look…
2 Likes
@priort could you share a video link about this process? It spiked my interest and I wanna see how this could be useful.
I could try but the process is essentially like copying your edits from one image to another. Instead you copy it to an identity lut and just save that. Now any software that supports luts can apply that to images…does that make sense?
1 Like
Ohhh ok. I think it does make sense now.
I basically do everything to the LUT that I did to the raw file (except sharpening, denoise etc.) and export it and then import it as a lut on whatever image I wanna style?
Now I wonder what would happen if I export a clut with LR Mobile and the black orange look, one with DT and the style I made and then compare the two.
This would cut down a bunch of modules in DT for sure.
Where do I get the neutral LUT?
The LUT I provided earlier is the result from the preset video (except for the aforementioned stuff that would corrupt and introduce artefacts, e.g. sharpening, clarity, etc.)
Fair warning: don’t use the one I attached in that post, specifically because it was converted to jpeg during upload, and you’ll need a clean png, compression is a no-no in this matter.
A link has already been shared, additionally in the following webpage you’ll find three other variants called a Hald5, Hald8 and Hald15. They are the Neutrals to which one applies the edits, I used Hald8 as an example as that’s what I’m used to (because of other software I used in the past) but of course. 3D LUT Creator - Materials and LUTs
By the way, try to follow the steps I listed earlier in this post,
it should help you get a feeling of the capabilities with translating edits from one software to another by emulating the colour transitions thanks to the files you will generate.
1 Like